Saying that a 5 μm filter element removes all the particles above 5 μm does not hold up in engineering. Between the nominal figure on the element and the retention that it can actually achieve lies a whole layer of efficiency information. Mixing the two together is the most common source of dispute in the element selection and at the acceptance.
The three ways of rating say different things
The micron rating gives a particle size figure, normally the nominal filtration accuracy of the medium, in other words the nominal pore size class of the structure of the medium. It describes the physical characteristics of the medium itself; it is not a removal rate for the particles of every size.
The β ratio is an efficiency rating. It is defined as the ratio of the particle concentration on the feed side to the particle concentration on the permeate side at a given particle size: a β of 10 means a retention of 90% at that particle size, 100 means 99% and 1000 means 99.9%. The β ratio is normally stated separately for several particle sizes, and the reading has to be read together with the particle size; a single number on its own means nothing.
The cleanliness class is a third system, and it describes a different object. The NAS 1638 grading is the cleanliness class of an oil sample after the particle counting; it describes the condition of the oil, not the performance of the element itself. An oil sample that reaches NAS 1638 class 5 after the treatment shows that the total particle content of the oil is within the corresponding range, and that has nothing to do with which element was used.
The difference between the nominal accuracy and the actual efficiency
The pores of a medium are not sieve openings of one size; they form a pore size distribution. An element rated at a nominal 5 μm has a real distribution of pore sizes, and whether a particle is held depends on its particle size distribution, on its shape, on the structure of the medium and on the formation of the filter cake during the filtration.
A deep filtration medium has a further dynamic change: when it is first put into service the efficiency of the element is not necessarily at its highest, and as the contaminants deposit inside and form a filter cake the actual retention changes. So with the elements from different batches, with the different oils and with the different running times, the same element behaves differently.
In other words, the nominal accuracy answers what pore size class the medium is, and the efficiency data answers how it actually retains. The first is easy to state and to compare, the second has to be read at a stated particle size and under stated test conditions. If the purchase document states only one figure in microns, the two sides can never have the same expectation of the result from the start.
Where the conversion most easily goes wrong
First, treating the nominal micron value as an absolute removal line, on the assumption that a 5 μm element holds back every particle larger than 5 μm. In fact the efficiency that corresponds to the nominal value has to be read off the efficiency curve of that medium, and the curves of the different materials differ a great deal.
Second, using the cleanliness class as the specification of the element when selecting. NAS 1638 class 5 describes the condition of the oil sample; it cannot be used to work out which element is needed.
Third, converting across the systems. There is no general conversion formula between the microns, the β ratio and the cleanliness numbers; the reference tables that circulate are all built on a specific medium, a specific particle shape and a specific counting method, and they can only be used as a rough guide. Writing technical requirements from a reference table carries a high risk.
| Rating | What it describes | What the reading means | Typical misuse |
|---|---|---|---|
| Nominal micron value | The structure of the medium | The nominal pore size class | Taken as an absolute removal line |
| β ratio | The filtration efficiency of the medium | The multiple of the retention at a given particle size | Quoted without the particle size |
| Cleanliness class | The particle condition of the oil sample | The class that corresponds to the particle count in the oil | Used as an element parameter in the selection |
- When reading the parameters of an element, the β ratio has to be read together with the particle size, for example the β ratio at a stated particle size.
- When writing the requirements at the acceptance, state whether it is a requirement on the efficiency of the element or on the cleanliness of the oil sample.
- When the three systems have to be brought into line, use the particle count of the same batch of oil samples as the reference instead of copying a reference table.
- Write the cleanliness judgement in one system only, NAS 1638 or ISO 4406, and do not mix the two numbering systems in the same document.
In my view the accuracy is a function and not a number; compress the function into a single number and the cost of the communication is settled on the day of the acceptance.